To me, "decentralized" doesn't mean every user has to run their own node. It means it's an open network, anyone can join, and there's no centralized control / censorship. You can buy a domain and use Proton, Fastmail, or any other provider to have your own email. If you become unhappy with them, it's easy to move. We have open specs that allow every service to interoperate, so you don't need to know or care what provider your recipient is using.
I don't think this is a fair characterisation of the article. She argues that email is an example of something that is failing as a decentralised application: Google's power over email is too strong. This is very different from claiming that it isn't an example of a decentralised application.
Email and the web are decentralised applications. But they are not the email and the web of the 1990s, when their decentralisation was the killer feature that saw them beat out centralised networks. To try and live on a non-Google/non-Facebook web is like using Linux on the desktop in 2000: it's certainly possible, but it's a bit of a hairshirt.
Spam filtering is a a very difficult thing to do well and Google's implementation was very good, very advanced, and trained on huge amounts of data. Something a normal person could never do on their own server and those who could would be spending a lot of time training the blocker which those using Google never even needed to think of.
It is not about centralization but about one player winning and beating the competition in that case.
So the author seems to take "decentralized" here to mean "fully decentralized and peer-to-peer", which is indeed an unconventionally strong definition. But later on, they mention IRC, which is centralized by any definition... Actually, the author doesn't seem to be working from a strong definition at all and they seem to have given this very shallow thought and research.
I must correct you here on "Protocols like Mastodon". The protocol that underlies the Fediverse is the W3C ActivityPub Recommendation [0] and making that distinction is really important. Not doing so would encourage centralization in the same ways as discussed elsewhere in this thread for Email vs. Gmail.
Mastodon is just a fediverse app that was an early adopter of ActivityPub, and managed to become quite popular. Part of this due to a better productization, and the jumpstart this gave them offering the best features first. To many newcomers Fediverse === Mastodon, but - while understandable - this is massively shortselling the interop capabilities of the protocol.
What has already happened by Mastodon dominance is that they are driving the technological direction, but not necessarily in the best possible way and to the overall benefit of the Fediverse. They have a custom client API for instance that is becoming a de-facto standard, adopted by others out of convenience.
"To clarify what I mean by decentralised: applications whose main purpose is fulfilled as part of a network, where that network is not reliant on any preordained nodes. Decentralised applications are also known as “federated”. There is also a more specific term “distributed”, also known as “peer to peer”. I won’t cover the distinction with the more specific terms because it’s not relevant here. I use the word application in a looser sense, to also cover protocols and specifications."
Conversely, in the blockchain space anyone can participate in the blockchain. If a miner refuses to mine your transactions, that's profit that another miner can pick up. And if worst comes to worst, you can set up your own mining infrastructure and mine that transaction yourself - if the system is working correctly, you never need permission from someone else at all to participate. (there are some assumptions here, and if you choose to reject those assumptions as valid you can reasonably claim that blockchains aren't decentralized either)
I'm confused, because it seems like you agree with the author. She says
>> My statement that email can be used without involving Google is technically true, but anyone they blacklist will quickly realise that they’re largely locked out of the global email network.
She is stating that the problem with email isn't that it is an open network, but that it is effectively centralized by a few big players. It might as well be a centralized service. The protocol may be open but if you aren't using a major centralized service you're going to experience a lot more difficulty, including potentially being black listed by these big players. Just getting accidentally classified as spam on one of them and you will locked out. Not to mention dealing with all the spam if you roll your own.
> If you become unhappy with them, it's easy to move.
I think this is an absurd statement and you know it or you are lying to yourself. Sure, forwarding exists but that's not exactly moving. If you've ever tried moving email addresses you'd know how difficult this is. It's significantly harder than moving phone numbers, which also isn't that easy. Just think of every account that you now need to modify. Sure, you don't need to port every account, but there's more than you probably think.
Email is (at the server level) a peer to peer network, only tied to DNS/BGP as the only central points, with a global namespace. That can't be said for a lot of open standards.
But, yes, that doesn't mean you have to run your own node. However, it should include the possibility to do so.
i can't see how i would host my own mail server with my domain, but then send emails from that domain via someone else.
the only way i could see that working is to use the ISP/other service as a relay to forward my mails.
I don't like cryptocurrencies (those who are unbacked and resource hungry, I like the tech behind) but bringing payments in the software world and out of financial gatekeepers could exactly the missing piece of the puzzle to have even more working decentralised apps.
Recognising this doesn't help with criticising capitalism, I guess.
I don't care that gmail has the majority of the market. I care that Google can not simply break compatibility with SMTP/IMAP and force everyone into their version of it. And no, "it's hard to self-host" is not enough of an argument. There are plenty of smaller professional hosting companies and as long as there is an option this is enough to stop Google from monopolizing the market.
Likewise, I don't care that Matrix still doesn't have a perfectly polished client. As long as it lets me talk and have video calls with my family and friends without being forced into WhatsApp or iMessage or whatever closed solution is popular, I'm willing to deal with its inconveniences and shortcomings. Bugs can be fixed, privacy violations can't.
Crypto as well. For all the talk about "investment opportunities" in crypto and the "get rich quick" schemes, there are still a good number of people working on it that understand that the important thing is to have an alternative to the existing financial system.
The worst part of the article though is not that it presents a problem that does not exist, it's the proposed solution: political intervention and regulation always works out in favor of the status quo. If you really want to kill whatever grassroots movement or attempt at breaking away from the big players, just throw the political/bureaucratic apparatus on top of it.
[0]: https://medium.com/incerto/the-most-intolerant-wins-the-dict...
A bold statement needs strong evidence. "Always" is already debunked with one counter-example. A look at the history of the labor movement or that of anti-trust legislation proves you wrong. Former pushed for worker-friendly legislation out of a grassroots and labor party-based movement, achieving laws that are often more demanding of larger companies. And latter may have seen a lack of enforcement since the Reagan era, but is chiefly centered on making big companies pay.
Never mind, it doesn't matter because it is completely orthogonal to my point. My point is that regulation works by applying top-down, global uniform of rules. It mandates centralized executive power, which is the polar opposite of bottom-up/localist/fractal/decentralized development of new tech and its applications.
"Making big companies pay" does not contradict "favoring the status quo". All that big banks want now is to be able to regulate crypto, so that they can keep away any threat of a disruptive business coming out of it. Same applies to Big Media and it is insistence of controlling information under the pretense of "battling fake news". Big Telco pushed for a long time to regulate OTT services, and this is until today it is so difficult to decouple voice from data plans. Facebook/ Twitter welcome regulation, because it will make basically impossible for small fediverse providers to grow and compete with them.
Regulation favors the status quo, how much more evidence do you need?
2021 is just a bad time to make confident pronouncements about decentralization. The vast majority of the population is living in Facebook/WhatsApp/Discord/Visa&Mastercard land, sure.
But I'm a heavy Matrix user, it works great. I torrent all the time: it works great. I'm also on urbit, and it works great. I don't use Mastodon personally but I hear few complaints from people who do. Podcasts are fine, other applications of RSS are pretty crippled by bad defaults of only including a little bit of the blog post, but Substack at least doesn't do it that way, which has me using an RSS reader again. It's nice!
The big elephant is the cybercoins, and it's just a weird time to declare them a failure. It's like no one remembers 2017, this has practically been a carbon copy, just with more institutional buy-in and bigger numbers.
No one knows if cryptocurrency is going to fulfill the promise of becoming a global store of value, let alone if it will be Bitcoin or Ethereum or some dark horse: but it pretty much has to go from worth nothing to worth everything, and the middle stages of that are going to be weird.
The one I'm really betting on is Matrix, though. They have reliable income from the French and German governments, and plenty of motive to make a really good platform. I fully expect them to be as slick and featureful as Slack and Discord within a couple of years, and start to eat those company's lunch.
"Redecentralization" is harder than the first phase of building a green-field decentralized Internet, because there are entrenched powers with a head start. But there are some compelling advantages to decentralized technology, with means, motive, and opportunity, to build them out.
Yeah. Don't think it will age well. Let's check back in around 2031 and see what actually happened.
A specific user from a niche geek website, a rounding error, an anecdote.
I'm glad for those fighting against these walled garden overlords, but let's keep in mind we're talking about a side with thousands of people VS a side with billions of people. Numbers matter, and the fall of these giants is still far away.
If anything, the battle is starting from recent anti-competition and privacy concerns in the public sphere, which is a good thing, but the public remains unaware of these niche distributed tools.
We are the ones who give the network effect to those services.
I hope it does age poorly. I tried to make it clear that I’m a prolific user of decentralised applications, and that I really want them to succeed. I just pointed out my feelings on why they usually don’t, in the hope we can do something about it.
I think this part of the article, instead of just being the last four paragraphs, should have been almost all of it. As it is, you make some very strong claims with no supporting argument whatsoever:
"the solution has to be political"
"The current system of profit motives, one might call it a market, was designed to optimise the process of extracting, refining, and transforming physical resources."
"some like to idealise the aforementioned market as a free and unregulated system, the truth is that it optimises rather poorly under those circumstances, and needs heavy regulation to align profit motives in the direction of efficient processing and distribution of resources."
And this in a footnote:
"I’m not convinced we should [keep these profit motives at all]"
All of this needs a lot of justification.
Element has come far, but Slack and Discord have more solid product direction.
I bought a star ages back but haven't done anything with it, I curiously read their update emails but remain tickled/skeptical by the oddball Hoon semantics and design language. Don't get me wrong, all with a layer of hopeful bemusement at the spectacle of it.
I'm very curious as to your thoughts.
It... works. It feels like running Matrix, or any other chat application, so it's easy to forget that everyone you interact with is a peer. Everyone is the admin of their own group, if they have one, and if they don't, they can start one by clicking a button.
Tlon is doing hosting now, or if someone is capable of setting up a droplet or whatever it's pretty straightforward to do yourself. Right now the hard part is getting a planet, because of the gas fees, but there's some kind of rollup solution coming along to make that cheaper. A comet remains free, and always will be; probably not the right answer if you want to run a group, but to get your feet wet, it works just fine.
There's some neat stuff coming down the pipeline, but right now urbit is stable, performs well enough, and it's cozy and calm. No one really advertises their groups, but if you find some urbit people on Twitter (which is easy) they can point you in some good directions to get started.
I don't think it even needs to be that. Most people (in the west) are tired of all of these apps, and their data collection, they just don't have an application that just works. If Element had better branding (maybe focusing on the Matrix bridges), then people would jump on board immediately.
Granted it's been a year since I last tried Matrix/Element, but back then it didn't "just work" despite many on HN claiming it did. My experience was far worse than Signal, which honestly I haven't had many problems with and many here claim it doesn't "just work."[0] I hit plenty of bugs in WA too. I think the difference is that I notice the centralized apps fix their bugs faster and that speed is proportional to the size of the parent company. Thing is if it is resolved quickly people just forget the bugs exist (and we also remember bugs more for apps which we haven't been using for long). There's a strong cult like mentality for Matrix, on HN, and I just don't get it (I don't get the cult mentality of Signal either, but HN is fairly anti-Signal so I'm not sure that needs to be addressed but we can). I love the idea of decentralization, but I just haven't seen it pan out and it still looks like it has a long way to go. I hope you all are right, but I'm not going to hold my breath and I don't think "just better marketing" is what's needed. There's a lot more.
[0] Disclaimer, I am still a Signal user because I want to support encryption and privacy and I don't see Matrix there yet and moving ecosystems is difficult. I'll take whatever encrypted and private service I can get friends to use that also doesn't require any tech literacy (this is an extreme must because most of my friends and family don't know a monitor from a desktop).
The folks at https://www.beeper.com/ are pursuing exactly that angle, all based on Matrix.
Exactly. This has been a re-run of 2017, and that is plenty enough reason to declare them a failure.
What happened to all the 2017 projects that raised billions during the last hype cycle based on breathless promises of decentralized wonderlands? They delivered absolutely nothing that anyone is using in the real world. "DeFi" is bullshit leverage products for speculators. NFTs are cryptokitties but somehow more stupid.
The whole scene is just a Ponzi Construction Kit for the gullible and greedy.
No. With Compound, Aave and dydx you have DeFi-Dapps that allow you to borrow Dollars (I.e. stable coins pegged to the dollar) and earn interest between 2-15% APR (there are no savings accounts out there that will get you that rate, at least in Europe with negative interest rates). All apps are integrated in the Coinbase wallet. In contrast to speculation on rising Bitcoin or Ether this is a rather safe investment. It’s only still hampered by the Ethereum transaction fees which are currently really high (but not all of the time).
Furthermore, there are decentralized exchange Dapps that seem to have been going on well so far even through the current dump.
If I were you, I would invest a little bit of your savings in Ether and just wait what happens. It’s got a huge potential to disrupt the banking business and change finance forever.
BitTorrent is used in a lot of settings for data distribution and for software distribution. Saying it's just for pirating reveals the author's laziness. Several of these use-cases are listed in the wikipedia page: https://en.wikipedia.org/wiki/BitTorrent
As for crypto, I recommend people read about Bitcoin being used for the purposes of sending remittances to start.
You don't always need to take over the world to be succesful. Its fine to have a niche and just do that really well.
I followed your recommendation and this is what I found: bitcoin might, in theory, work for the purpose of sending remittances, but in practice it doesn't.
So, returning to the original question, can cryptocurrencies offer an innovative solution for migrant remittances to developing countries? Yes and no. In theory, cryptocurrencies can really provide an effective and economical channel for money transfers to help alleviate poverty through remittances. However, there are two principle problems. Firstly, coins such as Bitcoin remain too volatile, in fact ten times more volatile than major currencies (Baur & Dimpfl 2021). This means that if migrants were to use such channel they would be exposing themselves to extreme risks. But, more importantly users of bitcoin may find themselves victims of speculative attacks, as the cryptocurrencies’ value remains completely speculatively constructed. Secondly, as with international bank transfers, cryptocurrencies need a bank account to buy and sell it. As explained above, having a bank account is not a given and represents a high barrier to entry for many of the poorest migrants.
https://blogs.lse.ac.uk/humanrights/2021/04/29/cryptocurrenc...
How it's being used is a different problem, but it's certainly effective and popular.
Blizzard has been using it for years: https://wowpedia.fandom.com/wiki/Blizzard_Downloader
This company is new and is using it as a CDN: https://www.peer5.com/
Bittorrent is everywhere, just like Linux, and you mostly don't notice it.
https://docs.microsoft.com/en-us/windows/deployment/update/w...
In all honestly, I've had very similar thoughts. As someone who started with BITNET before the Internet, I always believed "the Net" was supposed to level the field, and we no longer have to be "consumers" but rather willing participants capable of innovation, on our own right. I know, how naive.
Where I disagree is in the solution. It's not political. Whenever you think about the Internet, you MUST think global. I've had the chance to visit around 35 countries, and I've seen first hand how governments don't work for people, for the most part.
It comes down to this: we (geeks and nerds) need to get out of our comfort zone and think in terms of user experience. We CAN create amazing things. But we must think about the people using these solutions. People use Dropbox because it's easy, not because it's good. Same for Gmail and other services. If we are not creating solutions that can be used by anyone, in an easy and straightforward fashion, we're not doing our job. If companies don't see value in what we are creating when it could help them tremendously (i.e. IPFS) then WE are doing something wrong.
I think it all comes down to UX.
Yet, I think one big reason why non-profit solutions are so unpopular is not primarily about marking, but about the terrible, truly terrible UX that most non-profit projects have.
As far is I can tell this has (mostly) nothing to do with constrained resources and just with the extremely technical perspectives of the maintainers.
Element is a laggy Discord clone. Signal is years behind other messengers in terms of features and general UI quality, Thunderbird is stock in the early 2000s, same as VLC, and I don't know a lot of serious artists that are using GIMP.
So, even though this is surely not the only reason why open source projects are not widely used, I think it's a huge contributing factor.
There are shining example s though, like Blender, Godot or Mastodon, that are truly open source and have a great set of features paired with nice UX and I don't think they're doing too badly :)
Discord, for example, has 450 full-time employees and is valued at $7 billion. I simply don't see how small non-profit teams, many of which are run primarily by volunteer labor, can compete with that in terms of UX. How many full-time designers or UX folks are working on GIMP or VLC?
It seems to be a lot of more modern programs seem to have worse UX, in my opinion (part of the problem is lack of documentation, too); not whether or not it is profit. At least in my experience, profit makes many things worse because they try to be mass-marketing and more trying to earn money, rather than being first the quality of the product. And also, they put too many fancy animations and stuff.
In life, nothing is perfect. For example, I can't use a CPU based on an open ISA as my daily driver. Does that mean that CPUs "don't work"? Of course not.
The fact that something is used for illegal purposes means that it doesn't work? Does LSD "not work"?
The fact that something is hard to set up or maintain mean that it doesn't work? Does Linux not work? Do airplanes not work?
Does the fact that something is an environmental disaster mean that it doesn't work? Do cars not work? Do highways not work?
Imagine you came up with a technically superior version of email, and it functioned technically perfectly, but you were unable to convince anyone else to use it. Sure, it would "work" technically in a narrow functional sense, but it wouldn't "work" in terms of doing what it was actually designed to do. My taking LSD isn't really affected by other people doing so or not: it's a private thing. But my sending an email is: it's a network interaction.
The author's claim is correct in the sense that, if you hold a party and no-one comes, the party didn't really work.
None of that is relevant to the people who use drug marketplaces or cars, but those are different people.
When I write in more "narrative" tone, it's hard for me to make sure that my grammar is right, and that I'm making myself understood.
In fact, I always try to write my comments in narrative form first, but many times the grammar comes out awkward, so I scrap that and write in a more dialogical form.
If you have any recommendations on how to improve this, they're welcome.
"Consumers don't care"
The concept in the real world is simple: "If I want something I go to store x and buy it". Translating that concept in the digital world is super easy with centralization:
- Want to watch a video: go to YouTube - Want to buy something: go on Amazon - Want to host large amounts of files, buy S3 space
Even podcast, a technology I love and have been using for is in danger of becoming more proprietary. If you look at promotional material podcasts you'll often find statements like "listen on Spotify or Apple Podcasts" as if they are the providers of podcast. But Apple doesn't even host podcasts (yet)
We could have all of these things without centralization but just imagine the sales pitch for a RSS/BitTorrent backed video service with an open API and multiple competing frontends.
And in the example of YouTube centralizing everything (hosting, distribution and presentation) has great benefits for them (YouTube) so why would they not do it?
Like having the video and music and ebook (and VR headset!) you've paid for (and therefore think it belongs to you) arbitrarily removed or rendered inoperable with no recourse. If that happened with a DVD it'd be considered theft. Most normal people just don't think that can happen, or assume that if it did the legal system would be on their side, but it has and it does.
Ultimately, while I'd like to see more use of decentralized systems and API standard systems and federated systems, it's more likely that the biggest problems with centralized systems will be addressed ultimately through regulation, once the legal system catches up.
Last mile was always in the hands of oligopolies, if not monopolies. In New York, Andrew Cuomo (currently governor) strong-armed the last mile carriers into blocking Usenet for "think of the children" reasons. So how was the peer to peer Usenet killed off? Monopoly big business and big government.
These things don't exist in a vacuum - big business and the government work together to kill off decentralization and work towards centralization. Cuomo is a Democrat, but someone like Mitch McConnell is no different.
> My statement that email can be used without involving Google
I run my own mail server, and have kept compliant with what they want in terms of SPF and whatnot. They have an influence, but if they made Gmail a complete walled garden it would be self-defeating at this point.
Linux ran for years via e-mail and patch - huge Linux companies were IPO'ing in the 1990s and Linus was still using e-mails and patches to keep the system together.
I also still communicate via IRC. I'm sure a decentralized Slack-like IRC could be done if anyone cared to do so.
Git is a good example - it's a thing that is decentralized newer than other things mentioned.
Decentralization works just fine for me. Guess what decentralized in 2019? IT being centered in San Francisco.
IBM (and the seven dwarves) dominated centralized computing until the MOS microcomputers started selling like hotcakes in the 1970s. I have an Android phone in my pocket, and have futzed around with LineageOS (although did not put it on my phone yet). Who knew in 1974 that people could walk around with phones/computers in their pocket that was many times more powerful that an IBM S/360, that the people would have the source code to and could modify at will? That's a decentralization you aren't even aware was a decentralization if you're too young to have heard of white shirted IBM technicians shuffling punch cards.
Clarion is NOT A BLOCKCHAIN BASED project, so hopefully you won't have a tendency to dismiss it outright.
From the Introduction: "Clarion aims to give everyone in the world the tools to broadcast their message to everyone who wants to hear their message without creating dependencies on centralized infrastructure. It will achieve this with a censorship resistant "friend to friend" network which will leverage the unused resources of your friends and family to distribute your content.
The ultimate goal is to provide the performance and reliability of a "centralized service" with the freedom and independence of a logically decentralized network. With the help of the Clarion community we can free our friends and family from the tyranny of Twitter, Facebook, YouTube, Amazon, Apple and Google and produce a social network free from manipulation and 3rd party dependence."
[0] https://github.com/Clarionos/clarion [1] https://moreequalanimals.com/posts/book-launch
Most people don't matter and yeah, they 'll always prefer the prepackaged, "curated", high sugar content. The role of the decentralized web is to provide an infinite-frontier alternative so that ideas can grow outside the pettiness and the boxes of the mainstream. And yes, they do work for that
There is an entry cost to building decentralized tech, and it’s way too high and confusing IMO for mass adoption. The “cloud” is the easiest way to get up and running for both consumers and developers. But I don’t think it means dApps don’t work.
The beautiful thing about decentralization is that it doesn’t have to be profit aligned - it should be significantly cheaper to compute on your machine. We just need a better user experience to cross that barrier. I don’t think a decentralized platform that can overthrow the cloud experience exists yet. But that’s the hot opportunity and why it’s still emerging tech.
It always blows my mind that 1mb is 500 pages worth of information. Why do we pay $10/mo to store information that can fit on a thumb drive?
Which may be true, but i didn't find the evidence presented particularly compelling, and the structure of the argument was confusing.
All I can gather is the author once worked in blockchain and is now burnt out by the monetary side.
Everything they describe such as Git and email work fine. A strong technology will have SaS offerings but also still be usable from a CLI perspective.
Article title is click bait.
IM may be the first category where an internet application first appeared in a modern, user-friendly, full-featured form in a proprietary implementation before there was a standardized implementation.
[2] https://trashserver.net/en/
You're gonna hear some saying it's all oh so secure. But be aware of tradeoffs [3].
Additionally, considering several hundred million people use XMPP on a daily basis (and even more use something that started out based on XMPP, e.g. whatsapp), saying that XMPP is so terrible that it killed everything is so weird that it would need some facts to back it up.
I mean that instant messaging was the first time (on the internet, at least) that centralized networks beat decentralised networks (whereas centralised competitors to email, the web, etc. had lost), and the legacy of that is why many subsequent network systems (that indeed have very little in common) have been built as centralised rather than decentralised. And I've gradually reached the conclusion that the reason the centralised networks won that one was not because they were uniquely good, but that XMPP was uniquely bad.
> Additionally, considering several hundred million people use XMPP on a daily basis (and even more use something that started out based on XMPP, e.g. whatsapp), saying that XMPP is so terrible that it killed everything is so weird that it would need some facts to back it up.
Actually I see the fact that modified forks of XMPP have been so much more successful than mainline XMPP as a demonstration that XMPP's design makes a lot of unforced errors.
The notion that the internet would decentralize information is completely wrong. The internet restored the client-server relationship from the mainframe era, with the real power being with the databases. The internet recentralized computing after the PC era.
The human organism is (IMO) an awesome demonstration of both centralization and decentralization. As is the little blue speck we call our home biome.
What I might accept is that proponents of both arguments ("we need more centralization so we can provide a better service" vs "we need more decentralization so we have the freedom to evolve better services") ideally come to realize that there's a yen and a yang to the whole thing. That there's a lack of healthy balance/harmony between the two is a pretty easy argument to make now days.
BitTorrent is great but I wouldn’t exactly call it mainstream. I do wish more sites used it for downloads though.
I haven’t sat down and researched this, but an argument could be made that inherently a centralized app, who is trusted, available and resilient, will always be superior. Any thoughts?
If for some reason you don't want to use Google's platform, you're 100% free of creating your own mail server. The crazy thing is that it will probably work out just fine. The mail client/server model is really popular. You can go ahead and DIY your email and use all the popular clients without issue (maybe except Gmail? idk).
And sure, if you somehow become famous in a way Google doesn't like, they can effectively cut you off from lots of other people by not federating with your email server. In practice, though, this is probably pretty uncommon.
People have mentioned email and the Web as examples; I'd also throw in podcasts. (None of these are "apps," per se, but they're all examples from the article, so.)
> An argument could be made that inherently a centralized app, who is trusted, available and resilient, will always be superior. Any thoughts?
Yes. :)
The article makes the point, in so many words, that a lot of decentralized services end up dominated by a small number of major players. I think this is hard to argue with; yes, there are other service providers competing with the big ones, but that doesn't disprove the point. The most recent figures I can find suggest that Gmail had 43% of the email "market" in 2020; that's not more than everyone else combined, but number two is Yahoo (!) at 26%, with third place Outlook (i.e., the email service Microsoft keeps changing the name of, not the client application) trailing at 6%. Many people who might have been bloggers a bit over a decade ago use Medium or, more likely, just Twitter. (I think Twitter did far more to kill the blog than the death of Google Reader, but that's a different topic.) We can go on.
Where I think the article gets it wrong isn't in its description of the problem, it's in its diagnosis. Partially. I think the author is right in saying that better tooling isn't the problem, but wrong in saying that the profit motive is in "ensuring decentrialized systems fail." The profit motive is, to borrow your apt phrase, in making centralized systems trusted, available, and resilient. GitHub, Google Workspace, and Slack took over as much of the business world as they did because they're not only easier to set up than doing it yourself, they gain the benefits of a tremendous userbase: they're highly available and resilient, they're well-understood, they have lots of integration plugins.
I'd love to see more decentralized/federated systems succeed, but many of them aren't just fiddly and quirky, they're proudly fiddly and quirky. If they're going to take off, we may need to focus less on where we think companies behind centralized services are deviously undermining their decentralized relatives, and focus more on why those services succeeded in the first place.
The world wide web (the author complains DNS is too hard...well you dont technically need a domain name)
Usenet (was mainstream back in the day)
IRC is semi-federated, kind of obscure now, but was a bigger deal back in the day.
The author dismisses email, but that seems like an obvious answer.
DNS (hierarchical distribution. Depends how you define distributed i suppose) - what's more mainstream than that?
A lot of large companies also run their own servers, consumers can use icloud, protonmail, yahoo,... or the one from their ISP.
The argument in the article is that GMail controls email. It is definitely not true. Email is still how people do business, and business email is not owned by Google.
A lot of older networked games were, probably some modern ones too. One player hosts the game (either picked automatically by the match making system, or by running a game server listed by a index server), and everyone else connects to them. Call of duty, counter strike, starcraft, etc.
As other people mentioned the internet, also texting and telephones, but the peers in this example are really big. I think some of the bank transfer stuff is also a "distributed" system with big peers.
FireChat (distributed chat over wifi) was reasonably mainstream in Hong Kong for awhile: https://www.cbc.ca/news/science/firechat-off-the-grid-messag...
Not necessarily. They will still need an open protocol, good documentation, etc. And even then, it may not be suitable for what you are doing (and yet, they try to use it anyways). Furthermore, it won't necessarily always be available; you cannot predict some things that will go wrong with it. And, you cannot really trust them for sure either, necessarily.
www
1) when my project gets PRs, how do I hear about this most of the time? Yep, email
2) Nothing is forcing anyone to use github as their canonical repo. For my project, our canonical repo is self-hosted, and we use hooks to mirror to github. This lets people familiar with github use their tools and workflow, but avoids us being particularly tied into the github way.
Github and git are very different things, which is often forgotten. Github centralizes ever more functionality that is in any way related to the entirety of the software development process.
Want to set up a popular open source project hosted on your own code forge? Good luck, without the network effects of github your task is way harder.
Luckily there are movements towards decentralization and breaking of this domination with https://forgefed.peers.community and https://fedeproxy.eu
They dominate because they're better for end users who don't mind giving up compatibility/freedom to implement their own client, which the vast majority of people don't want to do. Slack/Discord/WhatsApp are far better instant messaging/voice or video conferencing software than IRC.
I think we’re only a few years out before Google uses it’s gmail monopoly to forcefully change email.
I use all of them for different groups and conversations, and my experience so far is that none of these is obviously or significantly better than another for most text/image-based group or one-on-one messaging unless you have specific niche needs.
Cutting "big tech" to size (big "tech" being a business sector misnomer that reveals the fake reality we all agree to swim in) will instantly propel decentralized protocols to mainstream: Smaller commercial operators will be forced to interoperate; Any usability issues or missing functionality will be developed in no-time.
The tentative link to the "physical" economy and externalities towards the end of the post is quite aposite and solutions to its own problems will be increasingly entangled with "solving" the digital economy. Just take the pandemic as an example: it is not hard to imagine digital infrastructure playing a key role in quickly identifying and limiting the spread of pandemics yet there was huge and justified mistrust that the monopolies will abuse their role...
Markets and virtual constructs such as corporates receive their legitimacy and license to operate from institutions serving the collective. When those institutions fail -> markets fail -> the collective suffers.
I think that the "abusive relationship with platforms like Medium, Twitter, and Instagram" is mostly psychological.
I've been noodling this point from the political end for academic reasons of late.
Conclusion: society is an 80/20 problem.
- 80% geared toward stability (why I want actual companies with heft doing things like bridges and hardware &c). Socialized things.
- 20% geared toward innovation. Risk taking. Capitalism. Experimentation of the "Hold my beer" variety.
The tension between these two cannot and should not be resolved; rather, graciously supported.
I submit that TFA makes the point that the 80% bullies the 20% a bit too much. Government's role is arguably to be an honest referee, except that the 80% tends to have the loot in addition to the numbers.
I'd like to rally more around the FSF to do more to preserve and protect open culture. Takes leadership, business chops, and patience.
Google, Twitter and Facebook, store and link to data and are therefor responsible for all kinds of things, like keeping porn away from kids, securing data, dealing with privacy, verifying age, dealing with local laws, managing the needs of rights holders, moderating discussions, deciding if Donald trump should be banned or not and so on, all while paying big $ to store all this crap. They are put in an impossible position that is taking a big tole on all of them, and may even give them so many political enemies that they get broken up.
A smart designer of a Decentralized application could see not storing data in a centralized storage, as an advantage because it lets them side step all of these problems.
This doesn't sidestep the core problem of curation, though. It depends on the nature of the application, but just looking at ones that involve displaying user-submitted publicly: if you don't have a frontend with some manner of filtering and moderation, the app will quickly be overwhelmed with a flood of spam and illicit activity (ranging from copyright infringement to child abuse).
Things like IPFS work great for decentralized content storage if you don't need to provide any sort of curation or frontend. But if you do, you'll quickly find yourself running what reduces to a centralized platform. Or if you're providing a federated service of some kind, then your users will be the ones who find themselves in that situation.
I agree, and I'd love to see Our Decentralized Future happen -- though I worry that 'smart designer' might end up implying one of the following in practice:
(a) smart enough to stick to a design so limited, in terms of P2P communication, that it'll not be a threat to current Killer Apps.
(b) smart, to the point of omniscience, regarding what a judge may decide is the designer's problem.
Also getting politics into this is the exact opposite of decentralization.
People should instead be incentivized to centralize in their own local communities, i.e. their family, friends, coworkers, local businesses, community centers, etc... Accomplishing this will likely require effort on multiple fronts, from increased regulation over the tech sector, to creating products that empower communities to run these services themselves.
But the final proposal would destroy all that is good about the internet:
>>What we’re dealing with here is far removed from physical resources, however, and the system has not been adapted at all. Although some like to idealise the aforementioned market as a free and unregulated system, the truth is that it optimises rather poorly under those circumstances, and needs heavy regulation to align profit motives in the direction of efficient processing and distribution of resources. If we are to keep these profit motives at all,7 we need new regulation to align them toward creating software that better serves our society.
What is needed is heavy government involvement in another way: subsidizing the development of open-source and decentralized applications. If they are better than the products produced by for-profit firms, then consumers will adopt them voluntarily, and if they're not, we wouldn't want consumers to.
Leaving the market free and having publicly funded projects compete with what the market produces on its own instills accountability for those allocating the public funds, by forcing them to get results better than what the private sector is producing on its own in order to validate their judgment on what should get funded.
This is what many people, author included, get wrong.
'Success' needs another defenition with decentralization, (and OpenSource, but that's not what this article is about) than with proprietary apps.
The absolute amount of users of Bitcoin, matrix, email, IPFS, mastodon and so on, is growing. In a staggering rate. The relative amounts are poor, but that really does not make it a failure.
But even if numbers dwindle, they hardly matter for making it a success. IRC is a massive success, because it allowed thousands of people to communicate for decades. And it paved the way for how we all work today. It's not an accident that Slack i see often called 'IRC done right', or so.
Decentralized apps are free by their nature, hard to monetize, the author made this central to her piece. But that means you need another defenition of success!
Bitcoin is a success the moment one single unbanked person can use it to store wealth or pay a friend. XMPP is a success the moment just two people can use it to communicate. Email is a success as long as one person runs it's own server that can deliver mail. Matrix is a success the moment one surpressed idea can be spread to one other person. And so on.
My thinking was about as anti-platform or anti-monopoly as it can be since, well, forever. But learning how to build business over the last decade has really opened my eyes to the harder parts of being a human and the hurdles to bringing value to market. So, here I sit with the skills to build just about anything I can imagine, a fairly deep understanding of broad business topics, and a bankroll... I'm working for someone else because honestly, the LoE to build something is dwarfed by the LoE required to identify and understand customers, and bring anything to market WITH centralized platforms. Going "decentralized" would increase that cost dramatically.
People are out here paying a third of the cost-to-customer right into the pockets of app stores because the stores create a safe market for consumers. The decentralized version looks like millions of people downloading and installing random .exe files from the internet. Then self-centralizing on a few sites with cringy names like download.com. Nobody wants that. Humans copy-pasta like crazy and naturally centralize.
I'd expect any decentralized system that fails to provide some unified discovery experience will fail. And if that system is centralized then the whole thing may as well be centralized.
> there’s someone who stands to make a large profit (or avoid losing their large profit) from ensuring decentralised systems fail.
But you can't actually make decentralized systems fail. Email's still around despite Google's abuse of spam filtering. RSS is still around despite Google's manhandling. Git is still around despite Github. IRC is still around despite Discord.
None of the truly decentralized options are as popular as the centralized-ish alternatives, but that seems entirely inevitable in an internet where most users have near-zero technical proficiency and want someone to hold their hand, which is just not viable without commercialization.
I think just like the world needs both a Linux and a Mac OS, it also needs centralized and decentralized options and while the decentralized options will drive actual progress, lots of people will get rich profiting from centralized bastardizations.
However, a larger omission is the fact that the new wave of federated services based on ActivityPub, Mastodon, Pixelfed[1], et all are missing from the article. I think that these disparate applications addressing different use cases, but which can talk to each other are the future of a federated web. Yes, there are larger entities that already carved a big chunk of the userbase, yes there are incompatibilities, but they're just starting and I'm optimistic.
[1] I'm talking from the perspective of someone that's building something in this space.
I do believe the author ran into this problem, especially in the VC world, except the author likely pointed this out herself.
I work for a major L1 project and never ran into this problem directly.
I've mostly witnessed less well funded projects bend towards their binance masters' desire for yet another pumpable token.
If you run into this problem, then you should spend their money building the fundemental components under a permissive or free software licensee. After some sunk costs then maybe they'll accept a more open business plan.
As an aside, VRFs are really amazing cryptographic tools for building distributed systems: https://www.youtube.com/watch?v=VyiG_JNxvIw
Mail is only usable because of spam filters, the best of which are completely opaque, and paid by ads. The web is open, but the search index isn't, and it spends a significant amount of time fighting search indexing spam / farms of autogenerated, low-value crap. Social networks need moderation to not become porn or terrorism supporting platforms, and the incentives or scale aren't there in a fully decentralized model. Phone networks are super open and right now in the US there's a nontrivial number of people whose inbound calls are almost all spam. Etc
Does policing need to be centralized? I seem to remember having a better time visiting small special interest forums spread around the web than I've had with the kind of everyone-everything-social networks that are ubiquitous today. They were typically moderated by some 1-5 volunteers from the community. If you didn't like their policies or the moderators you could use a different forum. The option to "vote with your feet" was more prevalent.
Even in large scale networks I think self-trusted blocklists could work to limit noise without any central authority on what is acceptable. It wouldn't prevent users from watching porn or discussing terrorism on their own accord, I guess, but that seems to be a problem with the current mainstream social networks as well.
The author presents themselves as familiar with blockchain applications, but I believe the statement above has two incorrect assumptions.
For one, if a developer aims to reduce energy use, they can deploy on a PoS chain.
Also, not every decentralized application has a "phony pretense of decentralisation". Sure there are some applications like that, but decentralized exchanges for instance, could not work anywhere else other than on a blockchain.
The idea isn't unlike competition / anti-trust laws. But would it require making new ones?
Centralization can happen for technical reasons but 99% of the time is about power and money, where a big company benefits enormously from owning a walled garden.
It was always very political. Feudal-society-vs-democracy level of political.
> The idea isn't unlike competition / anti-trust laws. But would it require making new ones?
Good question. If anti-trust laws were applied to network protocols the same way they are currently applied to all sort of physical goods, it would be already illegal to create walled gardens based on closed protocols (e.g. instant messaging).
"Power" (control) is important for security and compliance and centralized software offers it.
Imagine your role is to select tools that the engineers in your organization will use to collaborate as they create and access org IP, or even customer data. You personally favor decentralized F/LOSS solutions; you don't control the whole world, but why not deploy decentralized solutions in the realm that you do control? You'll just need to tick some boxes: How will you comply with a request to purge a certain customer's data? How will you restrict a former employee's access to your IP? How will you detect a hackers attempt to exfiltrate data from one of your decentralized nodes? What if you must provide employee communications logs to comply with a court order?
Ticking these boxes doesn't preclude a p2p file dissemination, but they do require centralized controls, centralized logging, centralized backups, centralized monitoring.
Because decentralized arrangements prevent or complicate compliance, even people who support decentralization and have the budget and authority to do something about it are prevented from rolling it out.
Same idea with the promises of the early WWW, and now with cryptos, decentralisation, etc. I could not be more agree with the author, while being upset that it is a very dark conclusion.
All major countries have discovered that extensive regulatory regimes are necessary to keep the free market in check. It isn't at all unusual that, as the article states, tech would require similar regulations to promote fairness, openness, and competition.
The underlying problem is paying for things. So many of the successes in decentralization come directly from universities where cost wasn’t really a factor.
Think of the watering hole analogy, developers who want to collaborate will eventually find some common location to assemble. If the features of this location are superior, or it has more popular projects, more developers will assemble, in a positive feedback loop, eventually this site will become a large, popular place for people to collaborate -- free or not. This behavior use to happen long before commercial activity was on the internet, back in the pre-Web listserv/majordomo/bitnet days. Certain people would become powerful gatekeepers purely by the size of their audience and the positive feedback effects of having a large audience.
That makes the central proposition that some kind of regulation is needed somewhat troublesome. I mean, take a large open source project like Linux or FreeBSD. A small group of maintainers, or a single dictatorial maintainer with veto power, wields enormous control over API and features that are included. If the maintainer gods decide not to support your hardware and refuse patches, does the government have to step in because it puts you at a disadvantage?
Levels of centralization in all things is probably inevitable. It wouldn't surprise me at all if there isn't some mathematics in network theory about scaling and hierarchy that would probably show it's inevitable -- even if you start out with something completely flat and fully connected graph, and without unequal incentives.
Basically, people with large audiences have larger power, whether it's commercial power, or even just social popularity.
The hypothesis is: capitalism is a hostile environment for user centric (non-abusive) software. I think I tend to agree (despite also believing that as an industry we could do a lot better job of pushing licenses that protect user freedom and help isolate users from abuse back against the domineering capitalists that simply don't understand how software ecosystems work).
I don’t think the author sufficiently defended their hypothesis. I didn't see any discussion about how and why profit motives are misaligned. The piece is missing evidence that supports the jump from “look some big companies have dominated decentralized systems” to “capitalism creates profit motives that must yield the results we’ve seen”. How are we sure the same result wouldn't happen under other economic systems, for example? Thats’s not a defense of capitalism on my part, just a desire to explore this topic much deeper.
Maybe software needs to be designed to resist domination? Or maybe the ability to be dominated has made certain applications successful? Maybe we shouldn't run some software for profit? Maybe we simply need to make it illegal to use persistent identity information in internet-scale applications and data sets beyond authentication and security purposes? Maybe our browsers aren't doing enough to present an “opt in on every use” type of experience for users (I wonder why)?
I do agree there’s an uncomfortable political element to the discussion.
It is indeed political, but i see it more as a need for distributed systems governing bodies to improve rather than for regulation of centralized ones.
Dark markets apart, it has almost no practical use today. I've read some comments on people sending money cross borders using bitcoin, the author might have overlooked this. Anyways, without numbers it is mere speculations... What is interesting in his article is the fact that decentralized solutions are deviating from their original purpose, becoming less powerful, and centralized.
And each year you can do much more, you just have to admit you dont understand it first, then you can learn.
Scott Alexander uses the example of the US interstate highway system to defend the type of “centralization” which has been so crucial in humanity’s increasing standard of living [1]:
> I like Seeing Like A State as much as anyone else (though see the caveats in Part VII of my review for some criticisms). But it worries me that everyone analyzes the exact same three examples of the failures of top-down planning: Soviet collective farms, Brasilia, and Robert Moses. I’d like to propose some other case studies:
> 3. The interstate highway system: 1950s army bureaucrats with a Prussia fetish decided America needed its own equivalent of the Reichsautobahn. The federal government came up with a Robert-Moses-like plan to spend $114 billion over several decades to build a rectangular grid of numbered giant roads all up and down the country, literally paving over whatever was there before, all according to pre-agreed federal standards. The public had so little say in the process that they started hundreds of freeway revolts trying to organize to prevent freeways from being built through their cities; the government crushed these when it could, and relocated the freeways to less politically influential areas when it couldn't.
Put simply, centralization is in our DNA. It is fundamentally how humans have become the planet’s most fearsome apex predator. Nature is very fearsome, of course, but the power to shape nature via concentrating power into the hands of the most capable in a meritocratic fashion is even moreso.
It’s not surprising centralized services dominate their decentralized counterparts. Bitcoin, however, is interesting here because the S&P is actually down since 1970 as measured in gold [2] — rather astonishing, given the S&P tracks some of the most powerful (centralized) human organizations in existence. It appears despite our best efforts, sometimes elements which exist outside of our political and legal grip can outperform even the greatest man-made feats.
Satoshi Nakamoto used clever cryptography to replicate a precious metal in digital form, as an attempt to pull humans away from centralized government monies. That was 2009, though, and people have since gotten too caught up in the glittery financial rewards aspect of Bitcoin to appreciate it on this level.
> Bringing fresh competition against Chrome isn’t really an option either. Modern browser rendering engines are so complex (especially when you have to implement all of Chrome’s undocumented quirks too), that even Microsoft gave up trying. And if you’re thinking of making your own system that eschews the web standards, good luck trying to bootstrap the entire web’s worth of content
The Gemini protocol is getting there. I recommend Lagrange [3]. Drew Devault hosts some of his blog articles on Gemini.
[1]: https://astralcodexten.substack.com/p/contra-weyl-on-technoc...
conversely, centralized internet applications are not infrastructure for other decentralized applications but they only exist for themselves.
i don't know if the internet highway should be considered centralized (as the backbone is in the hands of a few big players) but the use of that highway is certainly decentralized. the question is rather, can we have a federated bus system, where independent companies integrate their bus timetables so you can conveniently transfer from one to the other, or is greyhound dominating the whole network, refusing to cooperate with competitors?
Every time some piece of infrastructure became critically important for the society, people used political power to regulate the companies providing the infrastructure. Happened with railroads, water and sewers, electricity distribution.
The modern big tech with it's social networks is very similar to these older networks.
Of course, running a server means you get constantly attacked, or at least bothered.
This is not true.
In other words, what Bitcoin may or may not be today does not define what it is going to be in 2, 5 or 10 years.
It reads like someone just lost a lot of money in they recent dip and is venting.
For the record, There are other projects in the space that use less energy than bitcoin, like iota: https://howtobuyiota.co.uk/articles/how-energy-efficient-iot... - I read this article a while ago and full disclaimer, yes I invested.
I can't remember why it didn't catch on. I think maybe my mistake was not using the word "blockchain", so it lost out on the inherent sex appeal of the word.
Payment blockchains and IoT on the other hand is where I fail to see the relevance.
Firstly, I wanted to say that the writer seems biased, and due to this they disregarded the technology of blockchain. They mixed up bitcoin and blockchain and the use cases are completely different things
As for IOTA being a scam. I honestly do not know, I liked their use case but perhaps I am uneducated...If you have any papers or links to back up your claims then please provide them, i'm open to having my opinion changed.
